For those with little time:
There is much hype and enormous pressure to adopt AI in almost every aspect of our lives, but particularly in business, academe, and law. It is proposed as inevitable – use this or lose a great deal forever. We’ve seen these promises and threats before, and we should learn from the past not to open our arms and say “welcome.” This post offers some examples from recent times, and from long ago.
Take the courage to keep hold of your consent in all circumstances, and in regard to these latest developments, to stand far away and examine them from a distance. Don’t automatically shake this newcomer’s hand: watch carefully for a while to learn what the substances, good and bad, are. And then proceed cautiously if at all.
We should reconsider how we have built our world up to now, find our basic values, and discern what low-tech solutions may already be available to us. This is not to say that we must solve all our problems at the same time, nor by hand. But I wonder whether what we are seeing is a repeat of a phenomenon in earlier years1 (a May 2023 blog), the idea that to solve this problem here, we must look over there.
For those with more time:
Timothy Snyder (The Road to Unfreedom: Russia, Europe, America) suggests that America, Europe, and Russia all used to be guided by their own versions of the “politics of inevitability,” where in America “…nature brought the market, which brought democracy, which brought happiness.” In Europe, “…history brought the nation, which learned from war that peace was good, and hence chose integration and prosperity.” In the Soviet Union “…nature permits technology, technology brings social change, social change causes revolution, revolution enacts utopia.” This theme collapsed recently, supplanted by the “politics of eternity,” which replaces the inevitably better future with “…one nation at the center of a cyclical story of victimhood.” I think that much of Donald Trump’s rhetoric falls within this last category also.
The politics and commercial hype of inevitability and eternity leave one with an unchangeable future, except as authorized by AI and its proponents. Salvation is to be found not in the liberal arts where humane principles may be found, but exclusively in STEM. Teckies are having difficulty teaching morality and ethics to Anthropic’s Frame. Frame seems to more easily learn about larceny and malevolence than morality and ethics, from studying our on-line information about ourselves. We are assured that there are good purposes for AI, and also great risk https://archive.is/20251107051231/https://www.wired.com/story/ai-black-box-interpretability-problem/: it is possible for AI to act with malevolence, and to be deliberately deceitful. Morality and ethics can be entirely missing, as with humans. It seems that AI, learning from us, learns how to be bad.
When considering the hype on AI we need also to keep in mind both the huge business pressures to go all-in for the sake of those businesses, the ecological dangers (enormous energy consumption causing more emissions and/or dedicating nuclear reactors to the purpose), the enormous land use, and the devastating water consumption. https://unu.edu/inweh/news/environmental-cost-of-AIs-Enrgy-use-carbon-water-and-land-footprints.
More people are becoming alarmed at the potential of artificial intelligence, as well as the climate crisis, to bring frightening problems to our present and future.
We should reconsider how we have built our world up to now, find our basic values, and discern what low-tech solutions may already be available to us. This is not to say that we must solve all our problems at the same time, nor by hand. But I wonder whether what we are seeing is a repeat of a phenomenon in earlier years1 (a May 2023 blog), the idea that to solve this problem here, we must look over there.
John Lorinc’s Dream States: Smart Cities, Technology, and the Pursuit of Urban Utopias critiques the Google Sidewalk Labs proposal at Toronto’s Quayside on Lake Ontario. Google proposed gathering a great deal of digital information and monitoring pedestrian and vehicle traffic and movement, to make for better living within a new coastal development of the city. His critique is that the desire to collect information and greatly digitize the to-be-developed community had few if any demonstrable benefits for those living there, but it would transfer to Google a great deal of information about people’s daily lives. (Toronto withdrew from the project.) From this very recent event he looks back at the history of the development of urban infrastructure, with particular attention to roads, gas lines, communication lines, water and sewer, and electricity.
During his review, he refers to another book, From Warfare to Welfare : Defense Intellectuals and Urban Problems in Cold War America by Jennifer S. Light. She writes that the idea, early forms, and potential uses of digital and analogue communications and calculations technology were available and considered long ago. Beginning in the 1940s but particularly in the 1950s and 1960s, military and space research technology and methods of problem-solving were recruited to deal with the problems of urban physical blight, crime and violence. She provides a great deal of detail about this.
The idea began with the anticipated need to disperse urban areas to make them less susceptible to air bombing. This model was quickly absorbed into the analogy of cities as communications structures, which then led to the ideas of using TV, radio, telephone, and cable communications to enable people who were distant physically and socially from the centres of power, to better communicate with schools, libraries, government offices, and politicians. It is encouraging and astonishing to realize that by the 1960s there were proposals to use cable TV to effect that better two-way communication.
As we left the COVID era (somewhat), we in the Toronto area of Canada reflected on how radical and new it seemed to have on-line elementary education and secondary education, along with universities, and employment, be carried on remotely. Yet the ideas and early technology for this were considered and available sixty years ago. Light records that in the U.S. this potential radical use of cable was delayed by a six-year freeze on cable expansion imposed by the Federal Communications Commission at the behest of the broadcast industry. By the time the freeze was lifted, corporations saw the benefits of cable primarily in entertainment. Using it to include the disadvantaged in more of the civic functions, was abandoned.
This makes me reconsider whether commercial activities such as AI should be depended upon to provide, or allowed to inhibit, socially beneficial activities.
Perhaps it would be better to have more government initiative, funding, and planning, perhaps working through commerce to everyone’s benefit, or perhaps going it alone, using primarily human brain power. Familiar are the arguments that the free market should pretty much determine matters, and that government should do only what the free market cannot or will not. But reconsidering these specific examples suggests that more government action with less high tech might be more beneficial in the long run. And reconsidering AI in light of the latest discoveries, including Anthropic’s development of Mythos and Frame https://www.bbc.com/news/articles/c932g3v3e13o, it might be wise to examine carefully who makes decisions about how human effort is replaced or aided by nearly ungovernable automated forces.
Light asserts that much of the interest in space-age methods to solve earthly problems were based on ideas of national security which morphed from protecting cities from air bombing, to making cities secure from unrest and violence both criminal and political. The people and organizations involved saw urban development as the future source of their funding and continued legitimacy.
Whatever the history of thought, federal funding for various military and space projects and associated academic efforts began to decline. It just didn’t work out. For example, experiments with satellite technology failed. After some trials, cities returned to simple aerial photography to map.
While satellites could identify degraded urban areas (physical blight perceived by infrared photography), only neighbourhoods, not individual buildings, could be identified; whereas lower-altitude photography could get better pictures close-up. (Today’s satellites might do better.) Also, the top-down management of the military could not be applied to urban citizens’ problems because grass-roots consultations were needed to understand the unrest.
All these — high technology and information-gathering in a city area, systems applications rather than local consultations, high-altitude problem solving, –are examples of trying to solve a problem here by looking at something over there. I think the AI movement fits into this paradigm.
It is probably well and good to use AI as an aid; but not good to base human decisions exclusively or primarily on AI conclusions, nor to trust that AI has indeed examined everything that requires consideration such as moral and ethical dimensions.
And we must carefully examine how we feed and care for AI, as it competes for energy, land space, and water with other needs of humans, creatures, and the environment. We know that simple profit is the major driving force behind the enormous push of AI, followed by whatever good may be achieved. And we know this is a dangerous product.
Let us be sure to try to solve our problems here before we look over there and develop more problems from there while trying to solve these.
For those with even more time:
To organize our efforts here, let us reconsider some more of the past.
Gal Beckerman (The Quiet Before: on the Unexpected Origins of Radical Ideas) examines what was going on (conversations, meetings, rallies) before major revolutions. He examines movements from the beginning of the twentieth century on, into very recent years, e.g. Black Lives Matter, Black Identity, Dream Defenders, in reaction to assaults on Blacks in the U.S. and the resultant calls for defunding the police in Milwaukee. He recounts how the Soviet-era prisoners’ Samizdat communicated by hand-written and hand- delivered small pieces of paper, and how BLM and others tended to use social media.
But all can be described by a metaphor he borrows from Hanna Arendt, “the table”. He says that some of us need a table of our own, away from the main tables at which sit the powerful and important majority. A table where we can sit and talk with a few. He recounts how some groups got away from Twitter and the other large tables, and use smaller, secure apps for more limited communications. Some of these apps even inhibit immediate responses to a post, affording some time to contemplate. Some are so restrictive that it’s like writing a letter to someone, giving them time to read the entire message, taking some time to think, and then writing a letter back. (I don’t know why Gmail isn’t considered sufficient to this task, but there you are.) Slow communication, both low- and high-tech, has had some profound benefits.
These benefits of slow, quiet communication lead me to another book, The Republic of Letters, by Marc Fumaroli, an account of certain people in Europe in the 15th through 18th centuries. They felt the need to communicate with their peers (intellectuals, well-to-do businesses, academics, scientists) to preserve and share knowledge; keep discussion and dialogue going, and carry out cooperative efforts (such as viewing the same part of the sky at the same hour and reporting findings to help establish longitude), through the fog of never-ending European wars. This book has great detail about the format and courtesies involved in these correspondences. I urge readers to examine this book themselves. (The book is part of Yale University Press’ efforts to disseminate translations of many books to enable scholars and others to be aware of thought in other parts of the world.) As we see in this and in Beckerman’s book, the table analogy is very helpful and should be kept in mind.
Conclusions:
Not all technology over there is as useful here as it claims. Just as Toronto had to walk away from Sidewalk because the information gathered would benefit Google but not necessarily the community; just as U.S. cities returned to normal air mapping and turned away from satellite mapping; so some recent revolutionary thinkers needed to turn away from mass media to communicate around a table of their own, and the correspondents in the republic of letters needed to stay at their own table to keep civilized thought going.
So do we. But there must also be a time to join the big tables here. Join the big tables. Don’t throw them out with some sort of revolution. But think your own thoughts (the recurrent theme of all my blog posts). Bring your attention from over there to here. Communicate with humans, not machines.
Let us not accept AI as the only future.
1https://uponreconsidering.blog/2023/05/30/looking-over-there-for-solutions-for-problems-here-part-1/
Upon reconsidering... 